[1] The feathers of modern birds are so intimately associated with flight that their evolutionary origin was long assumed to lie in the demands of aerial locomotion. [2] On this traditional view, the earliest feather-like structures were incipient airfoils, and natural selection elaborated them because even rudimentary surfaces conferred some aerodynamic advantage. [3] The discovery of feathered fossils belonging to lineages that could not have flown has largely dismantled this assumption. [4] Many such specimens bear filamentous or downy integumentary structures on bodies whose proportions preclude powered flight, indicating that feathers predate flight by a considerable interval.
[5] If feathers did not arise for flight, what selective pressures account for their appearance? [6] Two hypotheses have dominated discussion. [7] The first holds that filamentous structures originally served thermoregulation, trapping a layer of air that reduced heat loss; on this account the insulating covering of small active animals was later co-opted, or exapted, for flight. [8] The second proposes that the earliest feathers functioned in visual display—signaling to rivals or potential mates—and points to fossil evidence of patterning and to the elaborate ornamental plumage of many living species. [9] These explanations are frequently posed as competitors, but the opposition may be overdrawn.
[10] A structure that insulates need not have been selected solely for insulation, and one that signals need not be useless for retaining heat; a single filament can perform both offices at once. [11] Some researchers therefore treat thermoregulation and display as concurrent rather than alternative pressures, each contributing to the retention and elaboration of integumentary structures at different stages or in different lineages. [12] What unites the modern hypotheses, and separates them from the older orthodoxy, is the recognition that the function for which a trait is ultimately conspicuous need not be the function that first favored it.
[13] This recognition carries a methodological caution. [14] Because soft tissues fossilize rarely and imperfectly, inferences about the original function of early feathers rest heavily on structures whose preservation is fortuitous and whose interpretation is contested. [15] To read a present utility back into the past—to assume that because feathers now enable flight they must always have tended toward it—is precisely the error that the fossil record has exposed. [16] The history of feathers thus serves less as a settled narrative than as a cautionary instance of how readily function and origin can be conflated when a trait's present utility is vivid and its earliest role is not.
